Subtle biphasic relationship is observed between cell migration speed and CG scaffold strut modulus; pore-size-dependent variation in cell speed is not explained by the variation in scaffold strut flexural rigidity. (A) Tetrakaidecahedral unit cell model for CG scaffold. For a series of scaffolds with a constant relative density, those with larger pore sizes, d1 (left), exhibit struts that are longer and thicker than a scaffold with smaller pore sizes, d2 (right). The longer/thicker struts have a greater flexural rigidity (Es × I) than the shorter/thinner struts, and would deform less under a constant cell-applied traction force. If changes in strut Es × I explain the pore-size-dependent effect on cell motility, cell speed would be expected to decrease with increasing strut modulus. (B) Average speed of NR6 cells migrating in scaffolds with a constant microstructure (pore size 96 μm) but with varying strut modulus (over the same range as due to changes in strut Es × I) shown via box-and-whisker (left) and mean ± SE (right) plots. From left to right, N = 116, 188, 191, and 79, respectively. Distinct from the concept of decreasing cell speed with increasing modulus, as predicted if the pore-size-dependent effects on motility are due to changes in strut Es × I, a subtle biphasic relationship is seen between cell speed and substrate modulus. Statistical significance (*p < 0.005) is determined by pairwise Kolmogorov-Smirnov test for nonnormally distributed data sets.